Oecologia (1998) 113:222±230 Ó Springer-Verlag 1998

Margaret J. Schoeninger á Urszula T. Iwaniec Leanne T. Nash Ecological attributes recorded in stable isotope ratios of arboreal prosimian hair

Received: 18 April 1997 / Accepted: 11 August 1997

Abstract Carbon and nitrogen stable isotope ratios apparently re¯ect the greater variation in diet among were measured in hair samples from two species of individual prosimians compared to individual monkeys. Galago from Gedi Ruins National Monument in eastern The isotope data reported here are equivalent, on aver- Kenya and from Lepilemur leucopus from Beza Ma- age, to those reported for other arboreal species from hafaly Special Reserve in southern Madagascar. Forest similar forest habitats and with similar dietary habits. structure was generally similar in the two areas but av- This supports the use of such data for paleoecological erage rainfall was lower in Madagascar. Species average reconstruction of forest and woodland systems and diet d13C values varied with feeding height in the forest reconstruction of extinct primate populations and canopy and with average rainfall level as expected from species. reported variation in plant d13C values. G. garnettii, which feeds higher in the forest canopy, had less negative Key words Primates á Feeding ecology á Stable isotope d13C values than G. zanzibaricus, which spends more ratios time below 5 m. L. leucopus, from a drought-a‚icted forest, had less negative hair d13C values than the two galago species. The values within the Lepilemur sample Introduction showed a positive linear relation with percent depen- dence on a CAM tree species and with xeric conditions Ecological data on primate species permit the testing of within the species reserve. Nitrogen stable isotope ratios general evolutionary models for extinct and extant spe- varied with trophic level of feeding and with time spent cies (Fleagle 1988) and the identi®cation of aspects of feeding on leguminous plants. The insectivorous galagos human behavior that most endanger or protect extant had signi®cantly more positive d15N values than the species (Pulliam and Babbitt 1997). Arboreal primates folivorous L. leucopus. Within the Lepilemur sample, can be dicult to study when they are nocturnal or d15N values varied inversely with the percent of feeding inhabit dense forests, so we tested the e€ectiveness of time spent on leguminous plants. The range of d15N and animal d13Candd15N values1 as indirect monitors of d13C values in each of the prosimian species is larger habitat use and diet. Additionally, we assessed the gen- than reported for animals fed monotonous diets and for erality of previous results showing that animal d13C New World monkey species. The monkey species feed as values were correlated with the level of forest cover and groups of individuals whereas the prosimians have sol- d15N values were correlated with feeding on leguminous itary feeding habits. The ranges in the prosimian species plants (Schoeninger et al. 1997). The former has potential

M.J. Schoeninger (&) á U.T. Iwaniec1 Department of Anthropology, University of Wisconsin, 1 Madison, WI 53706, USA A d value is de®ned as: e-mail: mjschoen@facsta€.wisc.edu; Fax: (608) 265-4216 Rsample d 1 L.T. Nash ˆ Rstd Department of Anthropology,  1000& (read as per mil or parts per thousand) Arizona State University,  13 12 15 14 Tempe, AZ 85287-2402, USA where R is the isotope ratio ( C/ Cor N/ N) and the standard is the internationally recognized standard. The international stan- Present address: dard for carbon 13C/12C is Pee Dee Belemnite (PDB), a marine 1Osteoporosis Research Center, carbonate; the standard for nitrogen 15N/14N is atmospheric ni- 601 N. 30th St., Creighton University, Omaha, NE 68131, USA trogen or AIR. 223 for paleoecological reconstructions (e.g., Schoeninger Materials and methods 1995) and the latter for assessing primate diet choice (e.g., Ganzhorn 1988; Oftedal 1991). Stable isotopes and prosimians Natural abundance stable isotope ratios in plant and animal tissues are being used with increasing frequency Patterned distributions of carbon and nitrogen stable isotope ratios to investigate various aspects of plant and animal ecol- occurring in the biosphere serve as the basis for the stable isotope ogy (Schwarcz and Schoeninger 1991; Association of approach to ecological questions. Across all plants, there is a non- Applied Biologists 1994; Koch et al. 1994; Lajtha and overlapping distribution of d13C values between plants that utilize Michener 1994; Pate 1994) but with little application to the C4 photosynthetic pathway (largely tropical grasses) and other plants such as trees and herbs, the majority of which follow the C3 non-human primates. We maintain that primates are pathway (O'Leary 1981, 1988). Herbivore values re¯ect aspects of ideal subjects for the method because their behavioral their ecosystem because their tissues average the variation in d13C ecology varies widely. Some species forage in groups values of plants selected as diet items (DeNiro and Epstein 1978; Bada et al. 1990). The method has commonly been applied to while others are solitary and can be expected to di€er 13 questions involving the feeding on C3 versus C4 plants. d C values individually in diet items selected. Species vary in the faithfully represent the relative amounts of browse versus grass in height within single forest canopies that they occupy and diet and are useful for animal species where observational data are in the types of forest utilized, and can be expected to ambiguous or scanty (Tieszen et al. 1979). As such, the values in re¯ect established variations in plant d values. In addi- human bone collagen track the spread of maize (a C4 plant) agri- culture in the prehistoric New World (Vogel and van der Merwe tion, across species there is a wide range of diet types 13 1977; Schoeninger and Moore 1992). Similarly, the d C values in from pure herbivory through omnivory to insectivory tooth enamel of extinct grazing animals monitor the presence of C4 which can be traced isotopically. grasslands during the Tertiary (Cerling et al. 1993; Kingston et al. To this end, we analyzed hair in three species of ar- 1994; Morgan et al. 1994). boreal, forest-dwelling prosimians living in Kenya and Primates, in general, show little use of C4 plants (Milton 1987) and the Kenyan and Madagascar forests are comprised mainly of Madagascar which have been subjects of multi-season C3 plants with additional species that utilize the third photosyn- behavioral studies (Harcourt and Nash 1986; Nash thetic pathway (i.e., crassulacean acid metabolism or CAM; Kluge 1994, 1997). The level of canopy cover in the forests was and Ting 1978; O'Leary 1981, 1988). Even so, the d13C and d15N similar to that in two of three regions of central America values in the prosimians should monitor various ecological vari- 13 ables. G. zanzibaricus feeds lower in the canopy and should have and South America where hair d C values in four more negative d13C values, on average, than G. garnettii, as plant monkey species correlated with the density of canopy leaf d13C values vary with canopy height (Vogel 1978b; van der cover (Schoeninger et al. 1997). Prosimians have been Merwe and Medina 1989; Broadmeadow et al. 1992; Garten and separated, phylogenetically, from New World monkeys Taylor 1992), and terrestrial animals have more negative bone collagen d13C values than arboreal ones (van der Merwe and Me- since the late Eocene or early Oligocene epoch (Simons dina 1991; Ambrose and DeNiro 1986). Lepilemur was expected to 1995) and, as such, serve as a taxonomic check on the show less negative d13C values than the galagos because the Beza general nature of the stable isotope approach to studies Mahafaly Special Reserve in southern Madagascar had undergone in primate ecology. The present study focussed on two a severe drought for the 2 years prior to hair collection. The d13C galago species, Galago garnettii Otolemur garnettii values in C3 and CAM plants, in contrast to C4 plants (Marino and ˆ † McElroy 1991), become less negative under drought conditions and G. zanzibaricus Galagoides zanzibaricus that live (O'Leary 1981; Ting and Gibbs 1982; Garten and Taylor 1992). In ˆ † 15 sympatrically at Gedi Ruins National Monument in terms of animal d N values, individual Lepilemur vary in the Kenya and Lepilemur leucopus from Beza Mahafaly amount of time spent feeding on leguminous plants (Nash 1997). 15 Special Reserve in southern Madagascar. The two areas, Those individuals which spend more time should have lower d N values than those which spend less time, because leguminous characterized as lowland, dry forests, are similar to each plants, in general, are less positive than non-legumes (Virginia and other in terms of overall tree height, the layers within the Delwiche 1982; Shearer et al. 1983; Shearer and Kohl 1986). This canopy, and the height and thickness of the under- pattern identi®ed prehistoric domesticated beans in South America growth. The Kenyan forest is multistratal with a canopy (Hastorf and DeNiro 1985), and the people feeding on them in central America (DeNiro and Epstein 1981). The two galago spe- at 15±20 m, emergents to 25 m, and a thick understory cies are omnivorous (Harcourt and Nash 1986) and should have (Harcourt and Nash 1986); these variables were not more positive d15N values than Lepilemur which is herbivorous quanti®ed at the Madagascar forest (Nash 1994), al- (Nash 1994) re¯ecting the stepwise increase in d15N from producers though canopy use was recorded for all three species to top consumers (Minagawa and Wada 1984; Schoeninger and (Harcourt and Nash 1986; Nash 1997). Annual precip- DeNiro 1984). This ``trophic level e€ect'' has identi®ed feeding relationships within single ecosystems (Wada and Minagawa 1983; itation in the Kenyan forest is around 1,000 mm (Har- Peterson and Fry 1987; Hobson et al. 1997) and, possibly, trophic court and Nash 1986) and around 750 mm in the position in less controlled situations (Schoeninger 1985, 1995; Madagascar reserve (Nash 1994). Although they vary in Bocherens et al. 1994). Starvation increases endogenous nitrogen 15 average body size (Harcourt and Nash 1986; Nash turnover (Swick and Benevenga 1977) resulting in N enrichment in tissues with rapid turnover (e.g., muscle and bone collagen in 1994), all three prosimian species are less than 1.0 kg. growing animals: Hobson et al. 1993), but none of the prosimians Such average body sizes normally correlate with insec- showed signs of food stress and this was not expected to be a factor. tivorous frugivory across primate species (Kay 1975, The e€ect of the drought on Lepilemur d15N values could not be 15 1984; Kay and Hylander 1978) but Lepilemur is a predicted, as elevated animal d N values correlate with water conservation mechanisms in some (Schoeninger and DeNiro 1984; complete folivore (perhaps coprophagous: Hladik and Ambrose and DeNiro 1986; Heaton et al. 1986) but not all (Am- Charles-Dominique 1974; perhaps not: Russell 1977; brose and DeNiro 1986; Sealy et al. 1987; Vogel et al. 1990) situ- L.T. Nash, personal observation). ations subject to drought conditions (Cormie and Schwarcz 1996). 224

Field studies imal's diet (Minson et al. 1975; DeNiro and Epstein 1978, 1981; Vogel 1978a; Jones et al. 1981; Nakamura et al. 1982; Tieszen et al. All behavioral data come from focal individual follows of radio- 1983; White 1993) and the collection of hair is less invasive than for tracked subjects. At Gedi, focal follow data comprised 130 h on other tissues commonly analyzed (e.g., bone collagen). Hair does G. garnettii (two males, three females) and 178 h on G. zanzibaricus not resorb or turn over, thus a speci®c period of feeding is moni- (®ve males, four females). About two-thirds of the follows at Gedi tored by each segment of hair analyzed. Among the few species were done between dusk and 0100 hours and the remainder be- studied, monkeys lose hair in molts (Inagaki and Nigi 1988; Dietz et al. 1995; Isbell 1995) which occur relatively rapidly (4±6 weeks tween 0100 and dawn. Approximately 250 h of focal follows were 13 15 conducted at Beza on Lepilemur (equally divided across four males once during the year). Thus, hair d C and d N values probably and one female). All follows at Beza were done between dusk and correspond to diets taken during a limited portion of the year. Hair 2400 hours. During focal follows, scan samples of behavior were growth patterns are less well-known in prosimians. Lemur catta, taken at 5-min intervals recording the individual's substrate, ac- the ringtailed lemur, reportedly shows reduced hair growth during tivity and food (when eating). Percent of time at di€erent heights of the dry season which is a period of reduced energy expenditure and substrate comes from these scan samples. Due to diculties in intake (Pereira 1993). Such a pattern of hair growth should result in observing galago feeding, most of the information on diet for the hair d values that are weighted toward wet-season diet values. galagos comes from analysis of feces from followed subjects and Neither Lepilemur nor G. garnettii showed any signs of molt. other animals (Harcourt and Nash 1986). The information on diet G. zanzibaricus did in some, but not all, individuals. No obvious in Lepilemur comes from scans which showed that all feeding was pattern of seasonality was seen in the molt (L.T. Nash, personal on leaves, stems, or ¯owers. Also, of 69 Lepilemur fecal pellets observation). examined from followed subjects and other animals, seeds were We analyzed hair from nine individuals from each species. In- found in only one pellet. In addition, during focal Lepilemur fol- dividual hair samples were cleaned sequentially with water and with lows, throughout each 5-min interval, we noted whether or not acetone and dried at 90°C. Approximately 3 mg was weighed into particular foods were eaten during that interval (1/0 sampling; quartz tubes with excess cupric oxide, copper, and silver. The tubes Martin and Bateson 1993). The information on diet elements is the were sealed under vacuum and the samples combusted at 900°Cin percent of all 1/0 intervals when any feeding was seen devoted a mu‚e furnace for 2 h. Tubes were allowed to come to room to Euphorbia tirucalli (Euphorbiaceae) or to Tamarindus indica temperature. Carbon dioxide and nitrogen gas were puri®ed (Leguminosae). Additional details on methods can be found in sequentially, collected cryogenically on a glass vacuum line, and Harcourt and Nash (1986) and Nash (1997). analyzed on a Finnegan MAT 251. A glycine laboratory standard analyzed repetitively produced a standard deviation of 0.2& in Several animals of each species were trapped for marking and to 13 15 attach radio-transmitters for tracking. Trapping and handling of d C and 0.3& in d N. Intra-animal variation, determined from the galagos was accomplished in 1988±1989, following methods several separate samples of hair that were cleaned and prepared from the same animal, showed d13C values within 0.2& of each described for the genus (Charles-Dominique and Bearder 1979). 15 The animals were sedated with ketamine for handling. Most of the other in eight sets of repetitive samples and d N values within Lepilemur subjects were captured by blowgun darts using Telozol 0.6& of each other in six sets of repetitive samples. This replica- (A.H. Robbins, Richmond Va) following published methods (Le- bility is the same as that reported in previous studies (Schoeninger mos De Sa and Glander 1993). After darting, animals were caught et al. 1997). in a canvas sling as they fell. A few animals were pulled by (gloved) hand from their sleeping holes and then sedated with Telozol for handling. In both studies, animals were released at their trap sites Results within 4±6 h of capture and there were no fatalities due to trapping procedures. The work on Lepilemur was done under a protocol No signi®cant di€erences in either d13Corind15N values approved by the Institutional Animal Care and Use Committee at Arizona State University. The work on the galagos predated such a were found between the sexes and the data were pooled. requirement, but followed methods that would currently be approved. Carbon

Hair analysis The galagos were 1.8& more negative in d13C values (Table 1, Fig. 1), on average, than Lepilemur (signi®cant Hair was cut from the back or tail as close to the skin as possible with a pair of ®ne-tipped surgical scissors. Among the galagos, at the 0.001 level of probability; 25 df ) even though all some of the hair was clipped from the tail as a method of identi- three species feed arboreally and most tree species follow 13 15 fying individuals. Hair d C and d N re¯ect the values in the an- the C3 photosynthetic pathway (O'Leary 1988). The

Table 1 Diet and stable isotope ratios (weight and feeding time values are from Harcourt and Nash 1986, for Galago; from Nash 1994, 1997, for Lepilemur)

Weight Time feeding d15N (AIR) (&) d13C (PDB) (&) (kg) n x SD Range x SD Range

Galago 0.14 30% fruit 9 7.4 1.2 6.4, 9.4 )23.3 0.2 )23.6, )23.0 zanzibaricus 70% animal prey 0% leaves Galago 0.84 50% fruit 9 7.1 0.9 5.6, 9.0 )22.8 0.2 )23.1, )22.5 garnettii 50% animal prey 0% leaves Lepilemur 0.6 0% fruit 9 5.5 1.0 4.2, 7.3 )21.3 0.8 )22.4, )20.1 leucopus 0% animal prey 100% leaves 225

Fig. 1 d13C values plotted against d15N values in hair (mean ‹ SD) for two Galago species from Kenya, Lepilemur leucopus from Madagascar and, for comparison, laboratory-reared Sus scrofa.The Fig. 2 Observational data of feeding for four individual L. leucopus ten sows were fed a monotonous diet and represent the minimal plotted against hair d13C values. There is a positive linear relation with interanimal variation expected. The o€set in d13C values between the time spent feeding on Euphorbia tirucalli, a CAM tree species. One two Galago species correlates with their feeding position within the result of the response to water stress in CAM species is that leaf d13C canopy. The d13C values in the Lepilemur individuals correlates both values are less negative than normal (O'Leary 1988). The Lepilemur with the drought conditions in their habitat and with time spent study site in Madagascar had experienced severe drought for 2 years feeding on a CAM plant species. The d15N values among the species priortohaircollection correlate with trophic position in feeding and with time spent feeding on leguminous plants among Lepilemur individuals negative of the galago hair d13C values they do not overlap them. Within the Lepilemur sample there is a galagos are insectivore-frugivores but insectivory ap- tendency for individual d13C values to correlate with a pears unlikely to account for the di€erence between the reported vegetation gradient of more mesic to more xeric genera. Although insects contain signi®cant quantities of moving away from a seasonally ¯owing river (Sussman lipids, and lipids are depleted in 13C relative to carbo- 1991; Sussman and Rakotozafy 1994). If we eliminate hydrates and proteins (DeNiro and Epstein 1977), con- two individuals whose habitats are severely degraded trolled feeding experiments show no evidence for and include a Euphorbia stand, individual d13C values incorporation of dietary lipid carbon into an animal's approach a signi®cant correlation with the distance of proteinaceous tissues (Ambrose and Norr 1993; Tieszen the individual's home range from the seasonal river and Fagre 1993). Rather, it appears that a drought e€ect which forms the eastern border of the roughly rectan- on CAM and C3 plants in the Malagasy forest accounts gular reserve (Spearman's R 0:716; n 7, critical ˆ 13 ˆ for the di€erence between Galago and Lepilemur even value for P < 0:05 is R 0:714). Leaf d C values in C3 though the forest is super®cially quite similar to that in plants are less negativeˆ under drought conditions, an Kenya. Several Lepilemur individuals were observed e€ect exaggerated in xeric areas compared to mesic areas feeding on the leaves and branches of E. tirucalli, a tall, (see Table 2; Garten and Taylor 1992) and there is a clear spindly, succulent tree in both the hot/wet season and trend for Lepilemur individuals from more xeric ranges the cool/dry season (Nash 1997). Various studies indi- to have less negative d13C values. cate that E. tirucalli is a CAM species. Leaf d13C values The two galago species are signi®cantly di€erent from 13 are less negative than those of C3 plants (Bender 1971), each other in d C values (Figure 1; 0:02 level of the Kranz anatomy diagnostic of C4 plants is lacking probability, 16 df). G. zanzibaricus is 0.5& more (Webster et al. 1975) and Old World succulent species of negative than G. garnettii. The smaller G. zanzibaricus Euphorbia usually have CAM (Kluge and Ting 1978). depends to a greater extent on insects than does the CAM plant d13C values are determined by the amount larger G. garnettii but, as discussed above, this is not the of day versus night CO2 ®xation (O'Leary 1981; Ting causal variable. The di€erence between the two galago and Gibbs 1982) which, in turn, is determined by water species is of the same order of magnitude and in the availability (Kluge and Ting 1978; Ting and Gibbs same direction as that between leaves from the base (0± 1982). Some succulents switch to CAM from C3 when 5 m) and the mid-canopy portion in deciduous, tropical water-stressed and revert back to C3 when conditions forests (see Fig. 3; Broadmeadow et al. 1992). G. zanzi- improve (Ting and Rayder 1982). Of four Lepilemur baricus spends approximately 70% of its time below 5 m individuals for which we have focal data (see Fig. 2), the in the canopy whereas G. garnettii spends over 50% of two that spent time feeding on E. tirucalli have the least the time above 5 m (Harcourt and Nash 1986). The d13C negative hair d13C values in the total Lepilemur sample. values in leaves from di€erent heights in individual Further, the two individual Lepilemur which spent no forests are comparable between similar forest habitats time feeding on Euphorbia have the most negative values (see Table 3) and result from isotope e€ects (Farquhar in the sample. While these values approach the least et al. 1982). These e€ects are produced by various fac- 226 tors (Lajtha and Marshall 1994) including soil-respired Nitrogen 12 C-enriched CO2 at the canopy base (Schleser and Jayasekera 1985; Broadmeadow et al. 1992) and high The two omnivorous galago species di€er from each light levels at canopy tops (Yakir and Israeli 1995). The other by only 0.3& (insigni®cant statistically) although average di€erence between galagos occurs even though G. zanzibaricus, which includes more insects in its diet the two species do not completely restrict their move- than does G. garnettii, shows the more positive value. ments to one or the other position within the canopy. The two galago species are signi®cantly more positive in Lepilemur, which occupies the upper level of the canopy hair d15N values than the herbivore, Lepilemur, on av- in a drought-a‚icted forest, also ®ts the trend. A high erage (signi®cant at the 0.01 level, df 25) as expected light level at the canopy top combines with the drought based on the di€erence in trophic position.ˆ Yet, the e€ect in the latter genus. average di€erence between the two genera is only 1.6& whereas the o€set between trophic levels (Minagawa and 13 Wada 1984; Schoeninger and DeNiro 1984) and between Table 2 E€ects of water availability on C3 plant d C values. Data are from Garten and Taylor (1992), in which habitats labelled animals and their diets (DeNiro and Epstein 1981; Hare 15 `xeric' are those on hilltops and hillsides and `mesic' refers to valley et al. 1991) is commonly 3&. Elevated animal d N bottoms. The same plant species were represented in each habitat values correlate with water conservation mechanisms in some situations (Schoeninger and DeNiro 1984; Am- Annual Habitat rainfall brose and DeNiro 1986; Heaton et al. 1986; Sealy et al. (mm) Xeric Mesic 1987; Cormie and Schwarcz 1996) and it is possible that d13C (PDB) ‹ SE (&) d13C (PDB) ‹ SE (&) the Lepilemur d15N values are elevated as a result of water stress. Further controlled studies are needed in 500 )28.9 ‹ 0.2 )29.8 ‹ 0.2 800 )29.2 ‹ 0.1 )30.0 ‹ 0.2 this area. 1,000 )29.8 ‹ 0.1 )30.2 ‹ 0.2 Within the Lepilemur sample there is a negative cor- relation between hair d15N values and the amount of feeding time spent on T. indica (kily), a leguminous tree (Fig. 4). Plant d15N values vary primarily with soil nitrogen loss rates (Nadelho€er and Fry 1994) and N2 ®xation by plant species with symbiotic bacteria (Shearer and Kohl 1986; Bowman et al. 1996; Hogberg et al. 1996). Legumes normally show d15N values that are less positive than non-legumes (Fig. 5), with the size of the o€set indicating the amount of N2 ®xation (Vir- ginia and Delwiche 1982; Shearer et al. 1983; Shearer and Kohl 1986). Among the four Lepilemur there is a 3& range of variation associated with a maximum dif- ference of 20% in feeding time devoted to leguminous plants. This range encompasses the total range of vari- ation within our Lepilemur sample. Fig. 3 Average leaf and animal d13C values plotted against vertical The range of variation within each prosimian species position within forest canopies. Leaf data are within level averages is large compared with previously reported data on hair reported for a deciduous tropical forest in Trinidad (Broadmeadow d15N values in four New World monkey species. The et al. 1992). The prosimian species (this study) occupy di€erent levels range of values in the two galago species (3.4 and within their deciduous tropical forests in Kenya (Galago)and & Madagascar (Lepilemur). The primate species data show a similar 3.0&) and in Lepilemur (3.1&) is up to an order of magnitude of di€erence in d13C values and the same direction of magnitude larger than in four Cebus (0.2&), ®ve Ateles di€erence as reported for leaf d13Cvalues (0.6&), nine Alouatta (1.0&), and seven Brachyteles

13 Table 3 Leaf d C (PDB) varies a b c d by position in canopy and by Canopy Trinidad Venezuela (3,500 mm rain) Tennessee Bavaria the amount of canopy cover in level C plants Deciduous Semi- Evergreen Evergreen Deciduous Deciduous 3 tropical evergreen tropical tropical temperate temperate tropical podzol laterite

Top )24.7& )28.4& )27.7& )27.9& Mid )28.0& )28.8& )30.5& )30.4& )29.6& )28.8& Base )28.6& )31.3& )35.2& )35.4& )31.5& a Data from Broadmeadow et al. (1992) b Data from van der Merwe and Medina (1989) c Data from Garten and Taylor (1992) d Data from Vogel (1978b) 227

Fig. 4 Observational data of feeding plotted against hair d15Nvalues for four individual L. leucopus (this study) and the average for nine Alouatta palliata (Schoeninger et al. 1997). Within Lepilemur,the stable isotope data show an inverse correlation with time spent feeding on Tamarindus indica which is a leguminous tree species. Within Alouatta, the population is reported to spend more than 40% of its feeding time on leguminous plants (Glander 1979). Legumes can ®x 15 atmospheric N2, in which case their tissues have d N values lower 15 than those in other plants, on average Fig. 5 Average plant d N values reported for N2-®xing species and for non-®xing species (data redrawn from Virginia and Delwiche 1982; Shearer et al. 1983; Shearer and Kohl 1986). The number of species from each collection site is shown above each column. Within (1.1 ) (Schoeninger et al. 1997) as well as in ®ve each site, the average value for ®xing species is always lower than that & for non-®xing species although there is variation between sites. For the domestic Sus fed monotonous diets (1.4 ) (M.J. Sch- 15 & most part, the d N values of non-®xing species fall between 6& and oeninger, U.T Iwaniec and T. Crenshaw, unpublished 7& andthoseofN2-®xing species are less than 4&. The variation is data). Lepilemur, the prosimian for which we have focal due to di€erences in source nitrogen (biomass degradation, source feeding data, shows a range of variation of 14% to 36% rock, or rain in non-®xing species, and N2 ®xation in the others) and nitrogen loss mechanisms in soil (biomass burning, water removal, of time spent feeding on leguminous plants. Alouatta plant uptake) was also reported feeding on leguminous plants (Glander 1979), yet, the range of variation in hair d15N values within Alouatta (Schoeninger et al. 1997) is one- correlated with trophic position and time spent feeding third that of Lepilemur. In contrast to Lepilemur, Alou- on leguminous plants. In addition, the solitary feeding atta forages in groups where individuals often use the habits of the three species are re¯ected in the large range same routes for travel following each other in single ®le of variation within each species when compared with (Milton 1980) and presumably eating similar items. Le- other species that feed as a group. pilemur and the galagos, for which we have no focal Of signi®cance is the apparent generality of these feeding data, have solitary feeding habits and presum- ®ndings. We contend that general trends and processes ably vary individually in diet items as reported for Le- are re¯ected in the d13C and d15N values of arboreal pilemur. We conclude that the large ranges in the three primate tissues. Average d13C values in Galago hair are prosimian species are probably due to variation in diet very similar to those in two species of New World items taken by di€erent individuals and that this monkeys living in dry, deciduous forests of Costa Rica parameter is an indirect indicator of species foraging and Brazil (Table 4). L. leucopus from a drought- 15 13 strategy. As such, the species range in d N values re- a‚icted CAM/C3 forest has less negative d C values ¯ects feeding habits while the individual values indicate and two species of New World monkeys from a wetter, diet. closed-canopy C3 forest in Costa Rica have more neg- ative d13C values. The d15N values from the species are also compara- Discussion ble. Two disparate genera, Galago and Cebus (Table 5), the former a prosimian from Kenya and the latter an As expected, average d13C values of species were corre- anthropoid from Costa Rica, are both insectivore- lated with the drought in Madagascar and with feeding frugivores and have identical average d15N values. There position in the forest canopy. Average d15N values is also congruence in two folivorous species, L. leucopus 228

Table 4 d13C in hair correlates with forest type. [Data for New the enormous potential in this approach to primate World monkeys are standardized to PDB (from Schoeninger et al. ecology, speci®cally, and to animal ecology, in general. 1997)]

Tropical Old World New World Acknowledgements We thank Melinda Carter, Jim Moore, Michele forest type prosimians monkeys Morgan, and an annonymous reviewer for comments on previous (annual rainfall) d13C ‹ SD (n) d13C ‹ SD (n) versions of the manuscript. Irwin Ting and Marion O'Leary pro- vided helpful discussions regarding Euphorbia and CAM plants as Drought )21.3 ‹ 0.8& (9) did Jim Moore on feeding habits in monkeys and prosimians. deciduous The research was supported by the Wisconsin Alumni Research (<750 mm) Foundation (to M.J.S) and by the Center for Field Research, Deciduous )23.1 ‹ 0.4& (18) )23.5 ‹ 0.3& (19) Wenner-Gren Foundation, National Geographic Society, and (<1;500 mm) NIMH grant 1 RO3-MH35736-01 (to L.T.N.). Evergreen )24.8 ‹ 0.3& (9) (4,000 mm) References

15 Table 5 d N in hair correlates with diet [data from Schoeninger Ambrose SH, DeNiro MJ (1986) The isotopic ecology of East et al. (1997)] African mammals. Oecologia 69: 395±406 15 Ambrose SH, Norr L (1993) Experimental evidence for the rela- Body Diet d N tionship of carbon isotope ratios of whole diet and dietary weight (kg) (AIR) ‹ SD (n) protein to those of bone collagen and carbonate. In: Lambert JB, Grupe G (eds) Prehistoric human bone: archaeology at the Insectivorous molecular level. Springer, Berlin Heidelberg New York, pp 1±38 frugivores Association of Applied Biologists (1996) Stable isotopes and the Galago 0.8 50% animal 7.1 ‹ 0.9 (9) & integration of biological, ecological and geochemical processes. garnettii 50% fruit Association of Applied Biologists Annual Conference in asso- 0% leaves ciation with The Society for Experimental Biology and the Cebus 2.5 20% animal 7.0 ‹ 0.1& (4) British Ecological Society, University of Newcastle upon Tyne- capucinus 65% fruit Conference Proceedings 15% leaves Bada JL, Peterson RO, Schimmelmann A, Hedges REM (1990) Folivores Moose teeth as monitors of environmental isotopic parameters. Alouatta 5±7 0% animal 3.3 ‹ 0.4& (9) Oecologia 82: 102±106 palliata 31% fruit Bender MM (1971) Variations in the 13C and 12C ratios of plants 64% leaves in relation to the pathway of photosynthetic carbon dioxide (>40% ®xation. Phytochemistry 10: 1239±1244 legumes) Bocherens H, Fizet M, Mariotti A (1994) Diet, physiology, and ecology of fossil mammals as inferred from stable carbon and Lepilemur 0.6 0% animal 5.5 ‹ 1.0 (9) & nitrogen biogeochemistry: implications for Pleistocene bears. leucopus 0% fruit Palaeogeo Palaeoclimatol Palaeoecol 107: 213±225 100% leaves Bowman WD, Schardt JC, Schmidt SK (1996) Symbiotic N2-®x- (15±35% ation in alpine tundra: ecosystem input and variation in ®xation legumes) rates among communities. Oecologia 108: 345±350 Broadmeadow MSJ, Griths H, Maxwell C, Borland AM (1992) The carbon isotope ratio of plant organic material re¯ects and A. palliata when plant species data are taken into temporal and spatial variations in CO2 within tropical forest account (see Fig. 4). Both species are committed folivores formations in Trinidad. Oecologia 89: 435±441 Cerling T, Wang Y, Quade J (1993) Expansion of C4 ecosystems as based on tooth morphologies (Kay 1975) and gut mor- an indicator of global ecological change in the late Miocene. phologies (Chivers and Hladik 1980) but Lepilemur has Nature 361: 344±345 average d15N values almost 2& more positive than the Charles-Dominique P, Bearder SK (1979) Field studies of lorisid average for Alouatta. The population of Alouatta spends behavior: methodological aspects. In: Doyle GA, Martin RD (eds): The study of prosimian behavior. Academic Press, New over 40% of its feeding time on leguminous plants York, pp 567±629 (Glander 1981), whereas the extreme in Lepilemur is 36%. Chivers DJ, Hladik CM (1980) Morphology of the gastrointestinal The possibility that some Lepilemur d15N values are ele- tract in primates: comparisons with other mammals in relation vated in response to water stress requires further study. to diet. J Morphol 166: 337±386 In combination, these data strongly support the Cormie AB, Schwarcz HB (1996) E€ects of climate on deer bone 13 15 d15N and d13C: lack of precipitation e€ects on d15N for conclusion that hair d C and d N values of forest- animals consuming low amounts of C4 plants. Geochim dwelling primates accurately re¯ect general aspects of Cosmochim Acta 60: 4161±4166 species habitat utilization and nutritional ecology. As DeNiro MJ, Epstein S (1977) Mechanism of carbon isotope frac- such, the data suggest that the analysis of hair (or other tionation associated with lipid synthesis. Science 197: 261±263 DeNiro MJ, Epstein S (1978) In¯uence of diet on the distribution proteinaceous tissues) from museum specimens of forest- of carbon isotopes in animals. Geochim Cosmochim Acta 42: dwelling primates can be used in reconstructing general 495±506 aspects of forest paleoecology, species diet, and species DeNiro MJ, Epstein S (1981) In¯uence of diet on the distribution feeding patterns (solitary vs. group feeding) in extinct of nitrogen isotopes in animals. Geochim Cosmochim Acta 45: 341±351 populations and species. Further analyses of animal Dietz JM, Baker AJ, Allendorf TD (1995) Correlates of molt in tissues from well-studied ecosystems in which diet items Golden Lion Tamarins (Leontopithecus rosalia). Am J Primatol are also analyzed can only hasten the full realization of 36: 277±284 229

Farquhar GD, O'Leary MH, Berry JA (1982) On the relationship Kingston JD, Marino BD, Hill A (1994) Isotopic evidence for between carbon isotope discrimination and the intercellular Neogene hominid paleoenvironments in the Kenya Rift Valley. carbon dioxide concentration in leaves. Aust J Plant Physiol 9: Science 264: 955±959 121±137 Kluge M, Ting IP (1978) Crassulacean acid metabolism: analysis of Fleagle JG (1988) Primate adaptation and evolution. Academic an ecological adaptation. Springer, Berlin Heidelberg New York Press, San Diego Koch PL, Fogel ML, Tuross N (1994) Tracing diets of fossil ani- Ganzhorn JU (1988) Food partitioning among Malagasy primates. mals using stable isotopes. In: Lajtha K, Michner RH (eds) Oecologia 75: 436±450 Stable isotopes in ecology and environmental science. Black- Garten CTJ, Taylor GEJ (1992) Foliar d13C within a temperate well, Oxford, pp 63±92 deciduous forest: spatial, temporal, and species sources of Lajtha K, Marshall JD (1994) Sources of variation in the stable variation. Oecologia 90: 1±7 isotopic composition of plants. In: Lajtha K, Michener RH Glander KE (1979) Howling monkey feeding behavior and plant (eds) Stable isotopes in ecology environmental science. Black- secondary compounds: a study of strategies. In: Montgomery well, Berlin, pp 1±21 GG (eds) The ecology of arboreal folivores. Smithsonian In- Lajtha K, Michener RH (1994) Stable isotopes in ecology and stitution Press, Washington, DC, pp 561±574 environmental science. Blackwell, Berlin Glander KE (1981) Feeding patterns in mantled howling monkeys. Lemos De Sa RM, Glander KE (1993) Capture techniques and In: Kamil AC, Sargent TD (eds) Foraging behavior: ecological, morphometrics for the Woolly Spider monkey, or Muriqui ethological, and psychological approaches. Garland, New (Brachyteles arachnoides, E. Geo€roy 1806). Am J Primatol 29: York, pp 231±257 145±153 Harcourt CS, Nash LT (1986) Species di€erences in substrate use Marino BD, McElroy MB (1991) Isotopic composition of atmo- and diet between sympatric Galagos in two Kenyan coastal spheric CO2 inferred from carbon in C4 plant cellulose. Nature forests. Primates 27: 41±52 349: 127±131 Hare PE, Fogel ML, Sta€ord TW Jr, Mitchell AD, Hoering TC Martin P, Bateson P (1993) Measuring behavior, 2nd edn. Cam- (1991) The isotopic composition of carbon and nitrogen in bridge University Press, Cambridge, UK individual amino acids isolated from modern and fossil proteins. Merwe NJ van der, Medina E (1989) Photosynthesis and 13C/12C J Archaeol Sci 18: 277±292 ratios in Amazonian rain forests. Geochim Cosmochim Acta Hastorf CA, DeNiro MJ (1985) New isotopic method used to re- 53: 1091±1094 construct prehistoric plant production and cooking processes. Merwe NJ van der, Medina E (1991) The canopy e€ect, carbon Nature 315: 489±491 isotope ratios and foodwebs in Amazonia. J Archaeol Sci 18: Heaton THE, Vogel JC, Chevallerie G von la, Gollett G (1986) 249±259 Climatic in¯uence on the isotopic composition of bone nitro- Milton K (1980) The foraging strategy of howler monkeys: a study gen. Nature 322: 822±823 in primate economics. Columbia University Press, New York Hladik CM, Charles-Dominique P (1974) The behaviour and Milton K (1987) Primate diets and gut morphology: implications ecology of the sportive lemur (Lepilemur mustelinus) in relation for hominid evolution. In: Harris M, Ross EB (eds): Food and to its dietary peculiarities. In: Martin RD, Doyle GA, Walker evolution. Temple University Press, Philadelphia, pp 93±115 AC (eds): Prosimian biology. University of Pittsburgh Press, Minagawa M, Wada E (1984) Stepwise enrichment of 15N along Pittsburgh, pp 17±22 food chains: further evidence and the relation between d15N and Hobson KA, Alisauskas RT, Clark RG (1993) Stable-nitrogen animal age. Geochim Cosmochim Acta 48: 1135±1140 isotope enrichment in avian tissues due to fasting and nutri- Minson DJ, Ludlow MM, Troughton JH (1975) Di€erences in tional stress: implications for isotopic analyses of diet. Condor natural carbon isotope ratios of milk and hair from cattle 95: 388±394 grazing tropical and temperate pastures. Nature 256: 602 Hobson KA, Sease JL, Merrick RL, Piatt JF (1997) Investigating Morgan ME, Kingston JD, Marino BD (1994) Carbon isotopic trophic relationships of pinnipeds in Alaska and Washington evidence for the emergence of C4 plants in the Neogene from using stable isotope ratios of nitrogen and carbon. Mar Mamm Pakistan and Kenya. Nature 367: 162±165 Sci 13: 114±132 Nadelho€er KJ, Fry B (1994) Nitrogen isotope studies in forest Hogberg P, Hogbom L, Schinkel H, Hogberg M, Johannisson C, ecosystems. In: Lathja K, Michener RH (eds): Stable isotopes in Wallmark H (1996) 15N abundance of surface soils, roots and ecology and environmental science. Blackwell, Berlin, pp 22±44 mycorrhizas in pro®les of European forest soils. Oecologia 108: Nakamura K, Schoeller DA, Winkler FJ, Schmidt HL (1982) 207±214 Geographical variations in the carbon isotope composition of Inagaki H, Nigi H (1988) Annual changes in hair length of the the diet and hair in contemporary man. Biomed Mass Spectrom Japanese monkey (Macaca fuscata fuscata). Primates 29: 81± 9: 390±394 89 Nash LT (1994) Behavior and habitat use of Lepilemur at Beza Isbell LA (1995) Seasonal and social correlates of changes in hair, Mahafaly Special Reserve, Madagascar. Am J Primatol 33: 230 skin, and scrotal condition in vervet monkeys (Cercopithecus Nash LT (1997) Vertical clingers and sleepers: seasonal in¯uences aethiops) of Amboseli National Park, Kenya. Am J Primatol 36: on the activities and substrate use of Lepilemur leucopus at Beza 61±70 Mahafaly Special Reserve, Madagascar. Folia Primatol (in Jones RJ, Ludlow MM, Troughton JH, Blunt CG (1981) Changes press) in the natural carbon isotope ratios of the hair from steers O'Leary MH (1981) Carbon isotope fractionation in plants. Phy- fed diets of C4,C3and C4 species in sequence. Search 12: 85± tochemistry 20: 553±567 87 O'Leary MH (1988) Carbon isotopes in photosynthesis. Bioscience Kay RF (1975) The functional adaptations of primate molar teeth. 38: 328±336 Am J Phys Anthropol 43: 195±216 Oftedal OT (1991) The nutritional consequences of foraging in Kay RF (1984) On the use of anatomical features to infer foraging primates: the relationship of nutrient intakes to nutrient re- behavior in extinct primates. In: Rodman PS, Cant JGH (eds) quirements. Phil Trans R Soc Lond B 334: 161±170 Adaptations for foraging in nonhuman primates: contributions Pate FD (1994) Bone chemistry and paleodiet. J Archaeol Method to an organismal biology of prosimians, monkeys and apes. Theor 1: 161±209 Columbia University Press, New York, pp 21±53 Pereira ME (1993) Seasonal adjustment of growth rate and adult Kay RF, Hylander WL (1978) The dental structure of mammalian body weight in ringtailed lemurs. In: Kappeler PM, Ganzhorn folivores with special reference to primates and Phalangeroidea JU (eds): Lemur social systems and their ecological basis. (Marsupialia). In: Montgomery G (ed) The ecology of arboreal Plenum, New York, pp 205±221 folivores. Smithsonian Institution Press, Washington, DC, pp Peterson BJ, Fry B (1987) Stable isotopes in ecosystem studies. 173±191 Annu Rev Ecol Syst 18: 293±320 230

Pulliam HR, Babbitt B (1997) Science and the protection of en- Swick RW, Benevenga NJ (1977) Labile protein reserves and dangered species. Science 275: 499±500 protein turnover. J Dairy Sci 60: 505±515 Russell RJ (1977) The behavior, ecology, and environmental Tieszen LL, Fagre T (1993) E€ect of diet quality and composition physiology of a nocturnal primate, Lepilemur mustelinus on the isotopic composition of respiratory CO2, bone collagen, (Strepsirhini, Lemuriformes, Lepilemuridae). PhD dissertation, bioapatite, and soft tissues. In: Lambert JB, Grupe G (eds): Duke University, Durham, NC Prehistoric human bone: archaeology at the molecular level. Schleser GH, Jayasekera R (1985) d13C variations of leaves in Springer, Berlin Heidelberg New York, pp 121±156 forests as an indication of reassimilated CO2 from the soil. Tieszen LL, Hein D, Qvortrup SA, Troughton JH, Imbamba SK Oecologia 65: 536±542 (1979) Use of d13C values to determine vegetation selectivity in Schoeninger MJ (1985) Trophic level e€ects on 15N/14N and East African herbivores. Oecologica 37: 337±350 13C/12C ratios in bone collagen and strontium levels in bone Tieszen LL, Boutton TW, Tesdahl KG, Slade NA (1983) Frac- mineral. J Hum Evol 14: 515±525 tionation and turnover of stable carbon isotopes in animal tis- Schoeninger MJ (1995) Stable isotope studies in human evolution. sues: implications for d13C analysis of diet. Oecologia 57: 32± Evol Anthropol 4: 83±98 37 Schoeninger MJ, DeNiro MJ (1984) Nitrogen and carbon isotopic Ting IP, Gibbs M (1982) Crassulacean acid metabolism. Waverly, composition of bone collagen from marine and terrestrial ani- Baltimore mals. Geochim Cosmochim Acta 48: 625±639 Ting IP, Rayder L (1982) Regulation of C3 to CAM shifts. In: Ting Schoeninger MJ, Moore K (1992) Bone stable isotope studies in IP, Gibbs M (eds): Crassulacean Acid Metabolism. Waverly, archaeology. J World Prehist 6: 247±296 Baltimore, pp 193±207 Schoeninger MJ, Iwaniec UT, Glander KE (1997) Stable isotope Virginia RA, Delwiche CC (1982) Natural 15N abundance of ratios monitor diet and habitat use in New World monkeys. Am presumed N2-®xing and non-N2-®xing plants from selected J Phys Anthropol 103: 69±83 ecosystems. Oecologia 54: 317±325 Schwarcz H, Schoeninger MJ (1991) Stable isotope analyses in Vogel JC (1978a) Isotopic assessment of the dietary habits of human nutritional ecology. Yearb Phys Anthropol 34: 283±321 ungulates. S Afr J Sci 74: 298±301 Sealy JC, Merwe NJ van der, Lee Thorp JA, Lanham JL (1987) Vogel JC (1978b) Recycling of carbon in a forest environment. Nitrogen isotopic ecology in southern Africa: implications for Oecol Plant 13: 89±94 environmental and dietary tracing. Geochim Cosmochim Acta Vogel JC, Merwe NJ van der (1977) Isotopic evidence for early 51: 2707±2717 maize cultivation in New York state. Am Antiq 42: 238± Shearer G, Kohl DH (1986) N2-®xation in ®eld settings: estima- 242 tions based on natural 15N abundance. Aust J Plant Physiol 13: Vogel JC, Talma AS, Hall-Martin AJ, Viljoen PJ (1990) Carbon 699±756 and nitrogen isotopes in elephants. S Afr J Sci 86: 147±150 Shearer GB, Kohl DH, Virginia RA, Bryan BA, Skeeters JL, Wada E, Minagawa M (1983) Nitrogen isotope fractionation as a Nilsen ET, Shari® MR, Rundel PW (1983) Estimates of N2- clue to food chain dynamics in marine environments. Tracer 8: ®xation from variation in the natural abundance of 15N in 2±12 Sonoran Desert ecosystem. Oecologia 56: 365±373 Webster GL, Brown WV, Smith BN (1975) Systematics of photo- Simons EL (1995) Skulls and anterior teeth of Catopithecus (Pri- synthetic carbon ®xation pathways in Euphorbia. Taxon 24: 27± mates: Anthropoidea) from the Eocene and anthropoid origins. 33 Science 268: 1885±1888 White CD (1993) Isotopic determination of seasonality in diet and Sussman RW (1991) Demography and social organization of free- death from Nubian mummy hair. J Archaeol Sci 20: 657± ranging Lemur catta in the Beza Mahafaly Reserve, Madagas- 666 car. Am J Phys Anthropol 84: 43±58 Yakir D, Israeli Y (1995) Reduced solar irradiance e€ects on net Sussman RW, Rakotozafy A (1994) Plant diversity and structural primary productivity (NPP) and the d13C and d18O values in analysis of a tropical dry forest in southwestern Madagascar. plantations of Musa sp., Musaceae. Geochim Cosmochim Acta Biotropica 26: 241±254 59: 2149±2151